Powers of FreedomReframing Political Thought

Nikolas Rose

published by the press syndicate of the university of cambridgeThe Pitt Building, Trumpington Street, Cambridge, United Kingdom

cambridge university pressThe Edinburgh Building, Cambridge, CB2 2RU, United Kingdom

http://www.cup.cam.ac.uk40 West 20th Street, New York, NY 100114211, USA http://www.cup.org10 Stamford Road, Oakleigh, Melbourne 3166, Australia

Nikolas Rose 1999This book is in copyright. Subject to statutory exception and to the provisionsof relevant collective licensing agreements, no reproduction of any part maytake place without the written permission of Cambridge University Press.

First published 1999

Printed in the United Kingdom at the University Press, Cambridge

Typeset in Monotype Plantin 10/12 pt [wv]

A catalogue record for this book is available from the British Library

Library of Congress cataloguing in publication data

Rose, Nikolas S.Powers of Freedom: reframing political thought / Nikolas Rose.

p. cmPortions of this book draw on published and unpublished papers

Acknowledgements.Includes bibliographical references.ISBN 0 521 65075 5 (hb) ISBN 0 521 65905 11. Political sociology. 2. Liberty. 3. Community. 4. Foucault.

Michel Contributions in political science. I. Title.JA76.R645 1999320 .01 1 dc21 9840306 CIP

ISBN 0 521 65075 5 hardbackISBN 0 521 65905 1 paperback

Contents

Acknowledgements page ix

Introduction: reframing political thought 1

1 Governing 15

2 Freedom 61

3 The social 98

4 Advanced liberalism 137

5 Community 167

6 Numbers 197

7 Control 233

Conclusion: beyond government 274

Bibliography 285

Index 307

vii

1 Governing

How should one analyse political power? For much of the twentiethcentury in European social and political thought, answers to this ques-tion were dominated by the massive spectre of the state. Whilst politi-cal theory in the United States up through the 1950s and 1960s wasmore pluralist in its vision of political power, even there, by the 1970sand 1980s, analysts were advocating the adoption of a state-centredapproach. The modern state was analysed in terms of an apparentlyineluctable tendency to centralize, control, regulate and manage. Socialand political theorists drew attention to this expanding role of the state,discovered the hand of the state even where it appeared absent, critic-ized prevailing conceptions of political pluralism because they seemedto ignore the structuring role of the state. In short, they wanted to bringthe state back in to the analysis of modern society.1

Over the last fifteen years, however, many sociologists and politicalscientists have argued equally vigorously in the opposite direction. Theyhave tried to find ways of thinking about and investigating politicalpower which are not immediately structured in terms of the hegemonicrole of the state, which recognize, in different ways, that modern systemsof rule have depended upon a complex set of relations between stateand non-state authorities, upon infrastructural powers, upon networksof power, upon the activities of authorities who do not form part of theformal or informal state apparatus. One sign of this movement has beenthe emergence of governance as a new field of social and politicalanalysis. At its most general, the term governance is used as a kind ofcatch-all to refer to any strategy, tactic, process, procedure or pro-gramme for controlling, regulating, shaping, mastering or exercisingauthority over others in a nation, organization or locality. This wideusage is certainly consistent with the definition provided in the OxfordEnglish Dictionary, which gives examples of use going back to Middle

1 This was the title of the very influential work by Theda Skocpol and others(Evans, Rueschmeyer and Skocpol 1985).

15

Powers of freedom16

English. Governance directs attention to the nature, problems, means,actions, manners, techniques and objects by which actors place them-selves under the control, guidance, sway and mastery of others, or seekto place other actors, organizations, entities or events under their ownsway. Used in this way, governance seems a handy and compendiousway of pointing to a number of important questions for investigation.Thus texts proliferate on the governance of the universities, of the healthservice, of the environment, even of cyberspace. The term seems auseful substitute and analogue for regulation, administration, manage-ment and the like, precisely because it is not overly burdened with con-ceptual baggage. But in the more specialized literature on governance,two more specific themes are evoked.

The first is normative.2 Governance can be good or bad. Governancetends to be judged good when political strategies seek to minimize therole of the state, to encourage non-state mechanisms of regulation, toreduce the size of the political apparatus and civil service, to introducethe new public management, to change the role of politics in the man-agement of social and economic affairs. Good governance means lessgovernment, politicians exercising power by steering (setting policy)rather than rowing (delivering services). Organizations such as theWorld Bank have sought to specify good governance in terms of stra-tegies that purport to disperse power relations amongst a whole complexof public service, judicial system and independent auditors of publicfinances, coupled with respect for the law, human rights, pluralism anda free press. They urge political regimes seeking aid and loans to corre-spond to this normative image of governance, by privatizing state cor-porations, encouraging competition, markets and private enterprise,downsizing the political apparatus, splitting up functions and allottingas many as possible to non-state organizations, ensuring budgetary disci-pline and so forth.

This normative sense of governance links to a second, descriptive,sense. This new sociology of governance tries to characterize the pat-tern or structure that emerges as the resultant of the interactions of a

2 The classic text here was Reinventing Government: How the EntrepreneurialSpirit Is Transforming the Public Sector by David Osborne and Ted Gaebler(D. Osborne and Gaebler 1992). Reinventing government became theslogan for the assault on big government in the United States, and the titleof a series of publications of hearings of the United States Congress Com-mittee on Governmental Affairs in the 1990s, examining issues ranging fromthe use of information technology in government to the restructuring of thepublic sector to deliver more for less.

Governing 17

range of political actors of which the state is only one.3 Governancerefers to the outcome of all these interactions and interdependencies: theself-organizing networks that arise out of the interactions between a vari-ety of organizations and associations. It is argued that these are of par-ticular significance today because recent political strategies haveattempted to govern neither through centrally controlled bureaucracies(hierarchies) nor through competitive interactions between producersand consumers (markets), but through such self-organizing networks.4

Politics is seen as increasingly involving exchanges and relationsamongst a range of public, private and voluntary organizations, withoutclear sovereign authority. Terms like actor networks, self-regulatorymechanisms, trust, habits and conventions, gift relations and infor-mal obligations are utilized to describe the actual operation of the com-plex exchanges through which governance occurs.

At first sight, it seems that, whilst the approach to political power interms of government has little in common with normative uses of thenotion of governance, it shares much with this new sociology of govern-ance. It is critical of the analytic utility of the classical concepts of politi-cal sociology: state and market, public and private, and so forth. Itagrees that new concepts are required to investigate the exercise of pol-itical power beyond the state. It argues that strategies of political rule,from the earliest moments of the modern nation state, entailed complexand variable relations between the calculations and actions of thoseseeking to exercise rule over a territory, a population, a nation and amicrophysics of power acting at a capillary level within a multitude ofpractices of control that proliferate across a territory. Like the sociologyof governance, it does not to deny the existence of states understoodas political apparatuses and their associated devices and techniques ofrule. Nor does it ignore the potency of the juridical and constitutionaldoctrines of state, sovereignty, legitimacy or the specific characteristicsof the legal complex in the programming and exercise of rule. But itrejects the view that one must account for the political assemblages ofrule in terms of the philosophical and constitutional language of the nine-teenth century, or that one must underpin this misleading account witha theoretical infrastructure derived from nineteenth-century social and

3 The most developed attempts to conceptualize governance can be found inthe work of Jan Kooiman and his colleagues (Kooiman 1993). See alsoLeftwich 1994.

4 Rhodes (1994, 1995) has used the phrase the hollowing out of the State todescribe this.

Powers of freedom18

political theory which accords the state a quite illusory necessity, func-tionality and territorialization.

From the perspective of government, the place of the state in specificstrategies and practices for governing different zones, sectors or prob-lems becomes a question for empirical investigation. Nonetheless, what-ever arguments may have been made by the philosophers of liberalism,the key characteristic of actually existing liberalism, as it developedover the second half of the nineteenth century and the first seven dec-ades of the twentieth, was what Michel Foucault termed the govern-mentalization of the State.5 That is to say, the invention and assemblyof a whole array of technologies that connected up calculations and stra-tegies developed in political centres to those thousands of spatially scat-tered points where the constitutional, fiscal, organizational and judicialpowers of the state connect with endeavours to manage economic life,the health and habits of the population, the civility of the masses and soforth. This governmentalization has allowed the state to survive withincontemporary power relations; it is within the field of governmentalitythat one sees the continual attempts to define and redefine whichaspects of government are within the competence of the state and whichare not, what is and what is not political, what is public and what isprivate, and so forth.

From the perspective of government, however, these developmentsare not best understood in terms of an relentless augmentation of thepowers of a centralizing, controlling and regulating state. The thesis,inspired by Jurgen Habermas, that the state has increasingly colonizedthe lifeworld is misleading, not least because the very nature and mean-ing of state and lifeworld were transformed in this process.6 State insti-tutions certainly extend the scope of their operations and the depth oftheir penetration into the lives of their citizen subjects. But they do soby a complex set of strategies, utilizing and encouraging the new positiveknowledges of economy, sociality and the moral order, and harnessingalready existing micro-fields of power in order to link their governmentalobjectives with activities and events far distant in space and time.7 Theselinks between the political apparatus and the activities of governing areless stable and durable than often suggested: they are tenuous, revers-ible, heterogeneous, dependent upon a range of relatively autonomousknowledges, knowledgeable persons and technical possibilities.

5 Foucault 1991: 103.6 See, for example, John Keanes analysis of the extent to which we are moving

towards a totally administered society (Keane 1984, ch. 3).7 Cf. Stoler 1995: 28.

Governing 19

Despite these similarities, the ethos of analytics of governmentality isvery different from that of sociologies of governance. First, analyses ofgovernmentalities are empirical but not realist. They are not studies ofthe actual organization and operation of systems of rule, of the relationsthat obtain amongst political and other actors and organizations at locallevels and their connection into actor networks and the like. In thesenetworks, rule is, no doubt, exercised and experienced in manners thatare complex, contingent, locally variable and organized by no distinctlogic, although exactly how complex etc. they are would be a matter fora certain type of empirical investigation. But studies of governmentalityare not sociologies of rule. They are studies of a particular stratum ofknowing and acting. Of the emergence of particular regimes of truthconcerning the conduct of conduct, ways of speaking truth, personsauthorized to speak truths, ways of enacting truths and the costs of sodoing. Of the invention and assemblage of particular apparatuses anddevices for exercising power and intervening upon particular problems.They are concerned, that is to say, with the conditions of possibility andintelligibility for certain ways of seeking to act upon the conduct ofothers, or oneself, to achieve certain ends. And their role is diagnosticrather than descriptive: they seek an open and critical relation to stra-tegies for governing, attentive to their presuppositions, their assump-tions, their exclusions, their naiveties and their knaveries, their regimesof vision and their spots of blindness.

No doubt, at any one time, one can find a whole variety of differentmethods in play for acting upon others, linked in a whole variety ofways. Their variety, and their linkages, are empirical matters and worthyof study. But what distinguishes studies of government from histories ofadministration, historical sociologies of state formation and sociologiesof governance is their power to open a space for critical thought. Thisstems from their preoccupation with a distinctive family of questions,arising from a concern with our own present. How did it become pos-sible to make truths about persons, their conduct, the means of actionupon this and the reasons for such action? How did it become possibleto make these truths in these ways and in this geographical, temporaland existential space? How were these truths enacted and by whom, inwhat torsions and tensions with other truths, through what contests,struggles, alliances, briberies, blackmails, promises and threats? Whatrelations of seduction, domination, subordination, allegiance and dis-tinction were thus made possible? And, from the perspective of our ownconcerns, what is thus made intelligible in our present truths (in a cog-nitive sense, but also in a bodily sense, in the sense of our habitualmodes of being in the world and experiencing the world and ourselves

Powers of freedom20

in it, and the ways in which the space of possible actions in that worldhas been put together) what do our studies of governmentality makeamenable to our thought and action, in the sense of us being able tocount its cost and think of it being made otherwise?

Perspectivism, here, is thus partly a matter of introducing a criticalattitude towards those things that are given to our present experience asif they were timeless, natural, unquestionable: to stand against themaxims of ones time, against the spirit of ones age, against the currentof received wisdom. It is a matter of introducing a kind of awkwardnessinto the fabric of ones experience, of interrupting the fluency of thenarratives that encode that experience and making them stutter.8 Andthe use of history here is to that untimely end it is a matter of forminga connection or relation between a contemporary question and certainhistorical events, forming connections that vibrate or resonate, andhence introduce a difference, not only in the present, but also in thehistorical moments it connects up with and deploys. As Gilles Deleuzeputs it, thinking of Nietzsche, things and actions are already interpret-ations. So to interpret them is to interpret interpretations: in this way itis already to change things, to change life, the present and oneself.

Governmentality

To analyse political power through the analytics of governmentality is notto start from the apparently obvious historical or sociological question:what happened and why? It is to start by asking what authorities of varioussorts wanted to happen, in relation to problems defined how, in pursuit ofwhat objectives, through what strategies and techniques. Such investi-gations do not single out a sector of the real for investigation ideas ratherthan events, for example. But they adopt a particular point of view whichbrings certain questions into focus: that dimension of our history com-posed by the invention, contestation, operationalization and transform-ation of more or less rationalized schemes, programmes, techniques anddevices which seek to shape conduct so as to achieve certain ends. Theydistinguish between historically variable domains within which questionsof government have been posed: the ways in which certain aspects of the

8 Creative stuttering is what makes language grow from the middle, like grass;it is what makes language a rhizome instead of a tree, what puts language inperpetual disequilibrium (Deleuze, He stuttered, in Deleuze 1997: 111).As Deleuze puts it, this stuttering is the moment in which language is strainedto the limits that mark its outside, when it is engaged in digging under stories,cracking open opinions, reaching regions without memories, destroying thecoherence of the self .

Governing 21

conduct of persons, individually or collectively, have come to be prob-lematized at specific historical moments, the objects and concerns thatappear here, and the forces, events or authorities that have rendered themproblematic. They investigate the ways in which debates and strategiesconcerning the exercise of political power have delineated the properrelations between the activities of political rule and different zones,dimensions or aspects of this general field of conduct of conduct forexample, the ways in which the management of virtue and vice has beencontested and divided between theological, pedagogic, medical and pol-itical authorities, or the regulation of economic life has been disputedbetween attempts at political management and claims made for the vir-tues of a self-regulating market guided by its own invisible hand. Theyconcern themselves with the kinds of knowledge, the ideas and beliefsabout economy, society, authority, morality and subjectivity that haveengendered these problematizations and the strategies, tactics and pro-grammes of government.

Governing here should be understood nominalistically: it is neither aconcept nor a theory, but a perspective. For sociologists of governancesuch as Kooiman and his colleagues, the object of investigation is under-stood as an emergent pattern or order of a social system, arising out ofcomplex negotiations and exchanges between intermediate socialactors, groups, forces, organizations, public and semi-public institutionsin which state organizations are only one and not necessarily the mostsignificant amongst many others seeking to steer or manage theserelations.9 But the object of analytics of government is different. Thesestudies do not seek to describe a field of institutions, of structures, offunctional patterns or whatever. They try to diagnose an array of linesof thought, of will, of invention, of programmes and failures, of acts andcounter-acts. Far from unifying all under a general theory of govern-ment, studies undertaken from this perspective draw attention to theheterogeneity of authorities that have sought to govern conduct, theheterogeneity of strategies, devices, ends sought, the conflicts betweenthem, and the ways in which our present has been shaped by such con-flicts. Far from reducing all to politics, they draw attention to the com-plex and contingent histories of the problems around which politicalproblematizations come to form cholera epidemics, wars, riots, tech-nological change, the rise of new economic powers and so forth. Suchproblematizations may or may not be shaped by previous strategies ofgovernment. In any event, they do not speak for themselves. They mustalways be individuated and conceptualized in particular ways. Political

9 See in particular Kooiman 1993.

Powers of freedom22

thought is thus not auto-effective: in making thought technical, attemptsat governing are always limited by the conceptual and practical tools forthe regulation of conduct that are available, although they may use themin novel ways and inspire the invention of new techniques. Far fromhomogenizing discursive space, these studies show how the space ofgovernment is always shaped and intersected by other discourses,notably the veridical discourses of science and changing moral rhetoricsand ethical vocabularies, which have their own histories, apparatusesand problem spaces, and whose relation to problematics of governmentis not expression or causation but translation. And far from offering anew theory of power, studies of government offer a perspective whichbrings into sight a domain of questions to be asked and practices tobe analysed. In particular, they seek to interrogate the problems andproblematizations through which being has been shaped in a thinkableand manageable form, the sites and locales where these problemsformed and the authorities responsible for enunciating upon them, thetechniques and devices invented, the modes of authority and subject-ification engendered, and the telos of these ambitions and strategies.

It thus seems to me to be useful to regard the notions of governmentand governmentality as marking out, in the most general way, the fieldupon which one might locate all investigations of the modern operationsof power/knowledge. The mechanisms and strategies of discipline andnormalization that Foucault analysed so provocatively in Discipline andPunish10 may certainly be seen in these terms. The prisoner or theschoolchild or lunatic may be constrained and confined. Disciplinarytechniques may be embodied in an external regime of structured times,spaces, gazes and hierarchies. But discipline seeks to reshape the waysin which each individual, at some future point, will conduct him- orherself in a space of regulated freedom. Discipline is bound to theemergence and transformation of new knowledges of the human soul.And discipline is constitutively linked to the emergence of new ways ofthinking about the tasks of political rule in terms of the government ofthe conduct of the population, or at least of those sections and zoneswhich have forfeited their claims to be contractual subjects of law orhave not yet acquired that right criminals, paupers, lunatics, children.

Similarly, the technologies of bio-politics and the biologized state thatare discussed in the first volume of The History of Sexuality11 can be seenin governmental terms. They are strategies which recognize and actupon the positivity of the domains to be governed the factors affecting

10 Foucault 1977.11 Foucault 1979a.

Governing 23

rates of reproduction and population growth, the genetic make-up ofthe race and the like. They seek, with some notable exceptions, to actupon these domains by reshaping the conduct of those who inhabit themwithout interdicting their formal freedom to conduct their lives as theysee fit. And if, in Nazi Germany, the freedom to act, indeed the veryexistence, of some subjects had to be erased, this was in the name of agreater freedom of the Aryan people and their destiny. Here, withoutthe controls exercised by liberal concerns with limited government andindividual freedoms, the despotism of the state that is always an imma-nent presence in all governmentalities is manifest in all its bloody ration-ality.

Foucault was far from consistent in his own thinking about therelations between the different ways in which he analysed power. Hecertainly cautioned against conceiving of a chronology that went fromsovereignty a discontinuous exercise of power through display andspectacle, law as command, sanctions as negative and deductive todiscipline the continuous exercise of power through surveillance,individualization and normalization to governmentality focusing onmaximizing the forces of the population collectively and individually. Atone point he suggested that one could identify a triangle of sovereignty,discipline and governmentality. Elsewhere he argued that it was simplis-tic to see the societies of normalization of the nineteenth century asdisciplinary: rather, in such societies, life was taken in charge by theinterplay between the technologies of discipline focused on the individ-ual body and the technologies of bio-politics, which acted on thosebodies en masse, intervening in the making of life, the manner of living,in how to live.12 I am happy to leave textual analysis to others.13 Frommy point of view, it is most helpful to consider that, in the powerregimes that began to take shape in the liberal societies of the nineteenthcentury, the thematics of sovereignty, of discipline and of bio-power areall relocated within the field of governmentality. Each is reorganized inthe context of the general problematics of government, which concernsthe best way to exercise powers over conduct individually and en masseso as to secure the good of each and of all. It is not a question of a

12 See the first volume of The History of Sexuality (ibid.), and the discussion ofFoucaults 1976 lectures in Stoler 1995, especially pp. 808.

13 It is worth noting, however, that in an interview in 1984, six months beforehis death, Foucault says: I intend this concept of governmentality to coverthe whole range of practices that constitute, define, organize, and instrumen-talize the strategies that individuals in their freedom can use in dealing witheach other (Foucault 1997: 300). I return to this issue of government andfreedom in chapter 2.

Powers of freedom24

succession of forms, but of the ways in which the discovery of new prob-lems for government and the invention of new forms of government embraces, recodes, reshapes those that pre-exist them. Reciprocally,within new styles of government, older forms, for example the claims ofan elected parliament to have assumed the mantle of sovereignty byvirtue of representing the sovereign will of the people, find novel spacesof deployment and new points of support, and enter new dynamics ofantagonism and conflict.

Rationalities

Of course, even terror can be a calculated instrument of government, ascan naked violence. No complex analysis is required to count the costsin lies and lives of these ways of exercising power. But the claim of theanalytics of government is that the modern regimes for the conduct ofconduct that have taken shape in the West, and those strategies thatcontest them, are ineluctably drawn to rationalize themselves accordingto a value of truth. Does it make sense to interrogate strategies ofgovernment in relation to truth? Is not government almost by definitionthe realm of the pragmatics of the possible, the territory of the deal-makers, of corruption, or pork barrelling and the like? Perhaps. ButFoucaults own work on governmentality implied that one could identifyspecific political rationalizations emerging in precise sites and at specifichistorical moments, and underpinned by coherent systems of thought,and that one could also show how different kinds of calculations, stra-tegies and tactics were linked to each. Thus Foucault and his co-workersexamined European doctrines of police and argued that these embodiedcertain relatively coherent ways of understanding the tasks and objectsof rule, which were codified and rationalized in particular texts and werelinked to a range of regulatory practices which would be hard to under-stand otherwise. This secular science of police was articulated in theGerman-speaking parts of Europe, and also in the Italian states and inFrance, in the period from about 1650 to 1800. It saw police not asa negative activity concerned with the maintenance of order and theprevention of danger, but as a positive programme (close to our contem-porary notions of policy) based upon knowledge, which could act as thefoundation of the power and happiness of States (to quote from thetitle of von Justis text of 1760-1).14 Ian Hunter has suggested that the

14 For discussions, see Knemeyer 1980; Oestreich 1982, esp. ch. 9; Raeff 1983;and Pasquino 1991. Andrew (1989) has a useful discussion of the argumentsmade by eighteenth-century charitable societies in Britain that their efforts

Governing 25

confessional churches played a key role in the emergence of this notionof governable populations, developing a particular understanding oftheir role as one of reshaping the conduct of their subjects in the nameof spiritual purity, deploying a range of measures for their spiritual disci-plining, and thus uniting those in particular geographical areas in moralenclosures.15 For present purposes, however, what is significant is that,whether solely secular, solely spiritual or as a conflict ridden hybrid ofboth, political and religious authorities now understood their powersand obligations in terms of relatively formalized doctrines of rule whichmade it both necessary and legitimate for them to exercise a calculatedpower over the conduct of populations of individuals, omnes et singula-tim (of each and of all).16

Others have shown relatively consistent projects of rationalization atwork in nineteenth-century liberal debates about the need to limit thescope of government vis-a`-vis a principle of the market or of the rightsof the individual: rationalities that make it easier to understand theapparent conflicts between doctrines that seek to delimit political inter-ventions and the proliferation of laws, regulations and apparatuses ofgovernment.17 The same kinds of argument can be made about the NewLiberalism that developed in Britain, Australia and the United States inthe late nineteenth century in an attempt to transcend the apparentlyirreconcilable positions of the advocates of individualism on the onehand and collectivism on the other.18 And a certain attempt at rationaliz-ation is at work in the architects of the New Deal in the United States,despite the fact that it is undoubtedly also the case that laws, organiza-tions and devices that it invented were often ad hoc attempts to addressproblems of unemployment, surplus production, farmers bankruptcyand crises of banking.19

It is not only liberal forms of government that operate according to acertain rationality: however barbaric were the murderous strategies ofgovernment in Nazi Germany, they were not simply acts of irrational

were an essential part of a good national police to maintain and refine civilorder.

15 Hunter 1998.16 See also Foucaults discussion of the idea of political reason in Foucault

1981.17 E.g. Dean 1991.18 Useful accounts of the New Liberalism, although not formulated in these

terms, are Clarke 1978 and Collini 1979; the standard source is Freeden1978.

19 On the New Deal, see Eden 1989; S. Fraser and Gerstle 1989; Finegold andSkocpol 1995.

Powers of freedom26

brutality or the institutionalization of resentment. It is true, as DetlevPeukert has pointed out, that The social history of the Third Reich wasa rank and tangled undergrowth of Nazi projects of reorganization whichnever got put into practice, rivalries and jurisdictional wrangles betweendifferent state, semi-state and non-state organizations, Schweikian strat-agems by the oppressed directed against the overweening demands ofthe bureaucracy, individual and collective acts of freebooting enterpriseand clamours for privileges, if necessary at the expense of others, anddeliberate self-sacrificing resistance..20 Nonetheless, Nazism fusedtogether a number of distinct elements into a relatively systematic matrixof political thought and action: a eugenic, biologizing and statist racism,prioritizing the management of the population through interventionsupon the individual and collective body in order to control lineages,reproduction, health and hygiene; a pastoralized dream of the multipli-cation of practices for the disciplinary regulation of the body politic inthe name of the race; the instrumentalization of the micro-fascisms ofeveryday life of the band, the gang, the sect, the family; and aredeployment of an older thematic of race, blood and earth.21 And, aswe know, the actual power of Nazism as a mentality of government wasits capacity to render itself technical, to connect itself up with all mannerof technologies capable of implementing its nightmarish dreams intoeveryday existence.

One is not dealing here with a scientific discourse regulated by theapparatus of experiments, proofs, journals, peer review and so forth.Nonetheless, political rationalities are characterized by regularities.22

They have a distinctive moral form, in that they embody conceptions ofthe nature and scope of legitimate authority, the distribution of authorit-ies across different zones or spheres political, military, pedagogic, fam-ilial and the ideals or principles that should guide the exercise of auth-ority: freedom, justice, equality, responsibility, citizenship, autonomyand the like. They have an epistemological character in that, as we shallsee in detail later in this chapter, they are articulated in relation to someunderstanding of the spaces, persons, problems and objects to be gov-

20 Peukert 1987: 24. Other books I have found most helpful on the politicalrationalities that were at play in Nazi Germany are Mosse 1978 and Proctor1988.

21 I am paraphrasing Foucaults remarks in The History of Sexuality (Foucault1979a: 14950) and Deleuze and Guattaris comments on the relations ofmacro-fascism and micro-fascism in A Thousand Plateaus (Deleuze and Guat-tari 1987: 21415). Note that the issue here is one of rationalities in theplural, not of a specific rationality of modernity (cf. Bauman 1989).

22 I am drawing here upon Rose and Miller 1992: 1789.

Governing 27

erned. And they have a distinctive idiom or language. A certain elementof thought, that is to say, is involved in all projects of government. JohnMaynard Keynes, one of the great thinkers of the social liberalism thattook shape in the United Kingdom between the First and Second WorldWars, might have been thinking of himself when he wrote, in 1926, thenext step forward must come, not from political agitation or prematureexperiments, but from thought.23

The critiques of welfare that have flourished over the past fifty years,in the post-war writings of neo-liberals such as Hayek, through the UScritics of the New Deal and the War on Poverty and in contemporarypost-social political arguments from left and right, seek to rationalizegovernment in new ways. Such strategic attempts to rationalize theproblems of government have effects. For example, the various tacticsenacted by the British Conservative government under MargaretThatcher in the 1980s were not realizations of any philosophy whetherit was Keith Joseph reading Adam Smith or one of his advisers readingHayek. They were, rather, contingent lash-ups of thought and action,in which various problems of governing were resolved through drawingupon instruments and procedures that happened to be available, inwhich new ways of governing were invented in a rather ad hoc way,as practical attempts to think about and act upon specific problems inparticular locales, and various other existing techniques and practiceswere merely dressed up in new clothes. But, in the course of this pro-cess, a certain rationality, call it neo-liberalism, came to provide a wayof linking up these various tactics, integrating them in thought so thatthey appeared to partake in a coherent logic. And once they did so, oncea kind of rationality could be extracted from them, made to be translat-able with them, it could be redirected towards both them and otherthings, which could now be thought of in the same way as, forexample, in the various deployments of the notion of entrepreneurship.And such rationalities were then embodied in, or came to infuse, awhole variety of practices and assemblages for regulating economic life,medical care, welfare benefits, professional activity and so forth.

It is not that the thought of Hayek, Friedman or anyone else for thatmatter was realized in neo-liberalism. It is partly that government con-tinually seeks to give itself a form of truth establish a kind of ethicalbasis for its actions. We can be cynical about this without, I hope, think-ing that this is just legitimization or ideology. To govern, one could say,is to be condemned to seek an authority for ones authority. It is alsothat, in order to govern, one needs some intellectual technology for

23 Keynes 1926: 53. I discuss social liberalism in chapter 3.

Powers of freedom28

trying to work out what on earth one should do next which involvescriteria as to what one wants to do, what has succeeded in the past,what is the problem to be addressed and so forth. When studies ofgovernmentality speak of liberalism, of welfare, of neo-liberalism andthe like, it is in this sense that these terms should be understood: not asdesignating epochs, but as individuating a multiplicity of attempts torationalize the nature, means, ends, limits for the exercise of power andstyles of governing, the instruments, techniques and practices to whichthey become linked. The name merely individuates an assemblagewhich may have been in existence for a long time before it was named,and which may outlive its naming. But nonetheless, the naming is itselfa creative act: it assembles a new individuation of concepts, symptoms,moralities, languages; it confers a kind of mobile and transferablecharacter upon a multiplicity.24

At many times and places, and more or less consistently in Europeand the United States since at least the middle of the eighteenth century,those seeking to exercise power have sought to rationalize their auth-ority, and these projects of rationalization have a systematicity, a historyand an effectivity. Each such project or strategy of rationalization, in thename of the market, in the name of the social, in the name of the libertyof the individual, is a strategy to intervene, whether in thought or inreality, upon a set of messy, local, regional, practical, political and otherstruggles in order to rationalize them according to a certain principle.Political rationalities are discursive fields characterized by a sharedvocabulary within which disputes can be organized, by ethical principlesthat can communicate with one another, by mutually intelligibleexplanatory logics, by commonly accepted facts, by significant agree-ment on key political problems. Within this zone of intelligible contes-tation, different political forces infuse the various elements with distinctmeanings, link them within distinct thematics, and derive different con-clusions as to what should be done, by whom and how.

Intelligibility

It is possible to govern only within a certain regime of intelligibility togovern is to act under a certain description. Language is not secondaryto government; it is constitutive of it. Language not only makes acts ofgovernment describable; it also makes them possible. This emphasis onlanguage is not at all novel. In relation to the history of political thought,

24 Cf. Deleuze on the function of the proper name, in Deleuze and Parnet 1987:1203.

Governing 29

it is most impressively exemplified in the method of historical critiquedeveloped by Quentin Skinner and his colleagues. This consists in asurvey of the language employed in order to identify the shared conven-tions (the distinctions, concepts, assumptions, inferences and assert-ability warrants that are taken for granted in the course of the debate)which render [certain acts] problematic and give rise to the range ofsolutions in order to free ourselves from the conventions of our age.25

However, analytics of government are not primarily concerned withlanguage as a field of meaning, or with texts embodying authorial inten-tions which may be recovered and made intelligible in the appropriatehistorical context. They are concerned with knowledges, or regimes oftruth.26 That is to say, they are concerned with historical epistemolog-ies: the reconstruction of the epistemological field that allows for theproduction of what counts for knowledge at any given moment, andwhich accords salience to particular categories, divisions, classifications,relations and identities.27 But perhaps even this epistemologicalcharacterization is a little misleading, to the extent that it might implya somewhat calm and peaceful succession of bodies of knowledge. Forit is not a matter of words, of concepts, of epistemologies, but of a wholeregime of enunciation. That is to say, an agonistic field, traversed byconflicts over who can speak, according to what criteria of truth, fromwhat places, authorized in what ways, through what media machines,utilizing what forms of rhetoric, symbolism, persuasion, sanction orseduction. It is not so much a question of what a word or a textmeans of the meanings of terms such as community, culture,risk, social, civility, citizen and the like but of analysing theway a word or a book functions in connection with other things, whatit makes possible, the surfaces, networks and circuits around which itflows, the affects and passions that it mobilizes and through which it

25 Tully 1995: 35. A good introduction to the work of Skinner is the collectionof essays edited by James Tully (Tully 1988).

26 It should be noted that I part company with those who have traced a line inFoucaults own writings from archaeology through genealogy to ethics.Analytics of government are concerned with truth, with power and with sub-jectification. For further discussion of the role of language in the exercise ofpolitical power, see the work of William Connolly and Michael Shapiro(Connolly 1983, Shapiro 1984).

27 Mary Poovey, from whom this quote is drawn, provides a good discussionof this in the context of social history (Poovey 1995: 3). The idea ofhistorical epistemology is derived from the work of French historians ofscientific discourse, notably Georges Canguilhem (Canguilhem 1994). Thepapers collected in T. Osborne and Rose (1998) provide a good introduc-tion to Canguilhems work. See also the discussion in Daston 1994.

Powers of freedom30

mobilizes. It is thus a matter of analysing what counts as truth, who hasthe power to define truth, the role of different authorities of truth, andthe epistemological, institutional and technical conditions for the pro-duction and circulation of truths.

As I have already suggested, one specific characteristic of modernstrategies of government is that they harness themselves to practices forthe production of particular styles of truth telling: the truth proceduresand pronouncements of objective, positive or scientific discourses what Georges Canguilhem terms veridical discourses. Thus the exer-cise of government has become enmeshed with regimes of truth con-cerning the objects, processes and persons governed economy, society,morality, psychology, pathology. Government has both fostered anddepended upon the vocation of experts of truth and the functioning oftheir concepts of normality and pathology, danger and risk, social orderand social control, and the judgements and devices which such conceptshave inhabited.

Perhaps there is a methodological point to be made here. A numberof historians of political ideas, notably J. G. A. Pocock and QuentinSkinner and his collaborators, have helped us understand that the writ-ings of philosophers Kant, Hume, Locke and others can beadequately understood only when located within the context of theproblems that concerned the political classes in their time and place,and the field of social and political discourse within which they wereengaged. They have also shown how many of those now assimilated intothe academic canon were actively involved in inventing and arguing fornew ways of governing, whether this be in the form of systems of tax-ation, practices of education or institutions for the reform of prisonersand paupers.28 There is no doubt that, in certain times and places, thearguments and activities of philosophers have played a significant rolein practices of government. This is a matter for empirical investigationin particular cases. But as significant, certainly since the middle of thenineteenth century, have been the truth claims articulated in texts thathave a less elevated character: statistical texts discussing the importanceof taking a census; proposals for reform of asylums written by doctorsof the insane; medical debates about the nature of cholera and themechanisms of its spread; economists arguments for the need for labourexchanges in order to create a true market in labour and to exposethe workshy and the unemployable for harsh intervention. This is notthe appliance of science, in the sense that the truths worked out in the

28 An excellent example is Tullys study of John Locke (Tully 1993); see alsothe work of Duncan Ivison, especially his book The Self at Liberty (1997b).

Governing 31

university study or the laboratory are being applied to specific practicalissues. Rather, it is in these pragmatic governmental arguments that theconcepts are forged that will later be formalized in theories, experi-ments, comparative studies and the like. And the methodological pointis this: it is, most often, at this vulgar, pragmatic, quotidian and minorlevel that one can see the languages and techniques being invented thatwill reshape understandings of the subjects and objects of government,and hence reshape the very presuppositions upon which governmentrests.

Governable spaces

Governing does not just act on a pre-existing thought world with itsnatural divisions. To govern is to cut experience in certain ways, todistribute attractions and repulsions, passions and fears across it, tobring new facets and forces, new intensities and relations into being.This is partly a matter of time. Thus, the invention of the factoryand work discipline involves novel ways of cutting up time in order togovern productive subjects: we must learn to count our lives by hours,minutes, seconds, the time of work and the time of leisure, the weekand the weekend, opening hours and closing time.29 The bell, the time-table, the whistle at the end of the shift manage time externally,disciplinarily. The beeping wrist watch, the courses in time managementand the like inscribe the particular temporalities into the comportmentof free citizens as a matter of their self-control.

It is also a matter of space, of the making up of governable spaces:populations, nations, societies, economies, classes, families, schools,factories, individuals. Mary Poovey and Peter Miller have used the termabstract spaces to characterize such governable zones.30 Poovey usedthe term abstract space in her examination of the different ways inwhich space is produced and organized in the exercise of power. She wasconcerned with the representational assumptions involved, for instance

29 Mariana Valverde pointed out to me the need to emphasize the role of tem-poralization in governing (see Valverde 1998c). This point is made classicallyby Edward Thompson in his analysis of time, work discipline and industrialcapitalism (E. P. Thompson 1967). It is also, of course, a key theme in Disci-pline and Punish (Foucault 1977).

30 Mary Poovey uses this term in her study of the construction of the socialbody (Poovey 1995); Peter Miller uses it in the context of an analysis ofthe fabrication of the domains upon which accounting and management willoperate (Miller 1994). See also Miller and OLearys analysis of the space ofthe factory (Miller and OLeary 1994).

Powers of freedom32

whether space is abstract or empty, the ways in which bodies arearranged in space, the metaphors of space such as machines or bodies,the relations of materialization or dematerialization that are involved.This notion of abstract space comes from Henri Lefebvre.31 ForLefebvre, space becomes abstract only as a result of the crushing of livedexperience and its vanquishing by concepts and representations. Butthese oppositions between the lived and the represented, the experi-enced and the conceptualized, the abstract and the concrete, seem tome to be misleading. Governable spaces are not fabricated counter toexperience; they make new kinds of experience possible, produce newmodes of perception, invest percepts with affects, with dangers andopportunities, with saliences and attractions. Through certain technicalmeans, an new way of seeing is constructed which will raise lived per-ceptions to the percept and lived affections to the affect.32 They aremodalities in which a real and material governable world is composed,terraformed and populated.

I think of these fabricated spaces as irreal. I take this term fromNelson Goodman.33 We need to use it with caution, lest we concedetoo much to the realists: reality is irreal; what else could it possible be?Goodman, in his foreword to Ways of Worldmaking says his irrealistposition is a radical relativism under rigorous restraints. It is a radicalrelativism because, for Goodman, there is no independent access to onetrue world against which our versions of it can be compared and evalu-ated. All we have are different versions of the world, versions con-structed out of words, numerals, pictures, sounds or symbols in variousmedia. We take particular versions for real largely as a matter of habit.From my point of view, however, Goodmans irrealism is too psycho-logical. His examples come from such well-worn and psychologicallycontentious claims as the relativity of the perception of colour, shapeand movement. His image of a version of the world is that of a picture.My own irrealism is technical, not psychologistic. It is technical in sofar as it asserts that thought constructs its irreal worlds through verymaterial procedures. Thought, that is to say, becomes real by harnessingitself to a practice of inscription, calculation and action.

One should not try to make up a theory of the fabrication of these

31 Notably in Lefebvre 1991.32 Gilles Deleuze is here talking about style in writing, music, painting; style in

governing may be more mundane but it is no less a process of creative fabul-ation (Deleuze 1994b: 170).

33 It is developed most clearly in his Ways of Worldmaking (Goodman 1978).Ian Hacking has discussed Goodmans work in a number of important papers(Hacking 1988, 1992).

Governing 33

irreal spaces: they are put together differently in different practices andcontexts. But we can make a few general points. The government of apopulation, a national economy, an enterprise, a family, a child or evenoneself becomes possible only through discursive mechanisms that rep-resent the domain to be governed as an intelligible field with specifiablelimits and particular characteristics, and whose component parts arelinked together in some more or less systematic manner by forces,attractions and coexistences.34 This is a matter of defining boundaries,rendering that within them visible, assembling information about thatwhich is included and devising techniques to mobilize the forces andentities thus revealed. For example, before one can seek to manage adomain such as an economy it is first necessary to conceptualize a setof bounded processes and relations as an economy which is amenableto management. One could write the genealogy of this economy, theways in which Adam Smith and David Ricardo presuppose that an econ-omy is more or less coincident with the territorial boundaries of a nationstate, and that trading relations between countries typically take placebetween distinct and relatively self-contained national economies.35 Itwas thus only in the nineteenth century that we can see the birth of alanguage of national economy as a domain with its own characteristicswhich could be spoken about and about which knowledge could begained. Once such an economy had been delineated, it could becomethe object and target of political programmes that would seek to evaluateand increase the power of nations by governing and managing the econ-omy. But spaces such as the economy are not brought into existenceby theory alone. For example, the strategies of national economic man-agement that were invented in the middle of the twentieth century weremade possible not merely by the installation of new sets of concepts tothink about the economy, but also through the construction of a vaststatistical apparatus through which this domain could be inscribed, vis-ualized, tabulated, modelled, calculated, national economies compared,indicators like rate of growth devised and so forth.36 And today, as the

34 I am drawing directly here on some formulations in Miller and Rose 1990.35 Barry Hindess provides a useful discussion of these issues in the context of

an analysis of contemporary arguments concerning the fragmentation of suchnational economic spaces (Hindess 1998a). Keith Tribe has written a com-pelling study of the development and transformation of notions of oeconomyand economy (Tribe 1978). The work of Donald Winch is particularlyinstructive on all these issues: see for example Winch 1969.

36 Grahame Thompson has analysed the forms of visualization out of whichknowledge and sense of the firm and the economy are constructed(G. Thompson 1998).

Powers of freedom34

discourse of globalization implies, this idea of the national economy isbeginning to fragment, and we can begin to analyse the shifting forces,conditions and forms of visibility that have allowed the deterritorializ-ation and reterritorialization of economic government, the emergence ofa novel conception of economic space.

Elsewhere I consider the formation of a number of governmentalfields or governable spaces. Here, however, I would like to make somegeneral remarks on ways in which one might analyse the spatializationof governmental thought.37 We can make a rough and ready distinctionbetween three dimensions or sets of problems.

Territorializing governmental thought

Governmental thought territorializes itself in different ways. As WilliamConnolly has pointed out, the term territory derives from terra land,earth, soil, nourishment but also perhaps from terrere to frighten,terrorize, exclude, warn off.38 It is a matter of marking out a territory inthought and inscribing it in the real, topographizing it, investing it withpowers, bounding it by exclusions, defining who or what can rightfullyenter. Central to modern governmental thought has been a territorializ-ation of national spaces: states, countries, populations, societies. Onecan trace the ways in which each of these territorializations takes shape.

Take, for example, society. One can chart how, perhaps beginningwith a book like Montesquieus Persian Letters published anonymouslyin Amsterdam in 1721, a certain way emerges for visualizing the formsof government, religion and habits of a particular country, includingones own, as one of a variety of forms in which human existence canbe organized in the shape of a society.39 Montesquieus sense of thevariety of laws, customs and usages of different peoples is undoubtedlyconditional upon the influx of information, speculation and inventionmade possible by the voyages of exploration, trade, missionary activityand colonization that Europeans had begun in the fifteenth century.40

This rendered the mores, habits and institutions that are taken forgranted by inhabitants of a particular locale simultaneously extraordi-nary and difficult to understand, and yet intelligible and explicable. This

37 I am drawing generally on Foucaults suggestions in The Birth of the Clinic(Foucault 1973) and more directly on joint work that I have undertaken withTom Osborne on the history of empirical social thought (T. Osborne andRose forthcoming).

38 Connolly 1995: xxii.39 Montesquieu [1721] 1971.40 Richter 1977: 32.